DEPARTMENT OF ASTRONOMY AND COSMICAL PHYSICS. 491 



reilection, aud the consequent diificulty in determining the nadir point. But this 

 is by no means the only disturbing effect. The object of the present note is to 

 recall attention specially to a subtle source of error discovered by Dr. Robinson in 

 an exaggerated form, which in some degree probably affects all observations for 

 position. He found that if after the bisection of a star, but before the micro- 

 scopes were read, a train was heard to pass, the shaking caused a displacement of 

 the instrument, and the record might consequently be erroneous by as much as 4". 



In his explanation Dr. Robinson goes into detail, and refers to the slow-motion 

 screw which is used to make the observation. His particular explanation would, 

 therefore, not always be suitable in the case of other instruments which may 

 have no such screw. At Greenwich, for instance, there is no screw connected 

 with the clamp, bisections of the star being made with the telescope micrometer. 

 But it seems probable that a general form can be given to the explanation to suit 

 all instruments. If an instrument be turned from one position into another very 

 different it will not take up its new position all at once, but for a few minutes 

 after it has been set there will be slow, minute changes going on, arising from the 

 yielding to stresses and the giving way of friction. This is no mere theoretical 

 supposition, for such changes have been proved to occur in the case of tlie 

 Greenwich transit circle by a long series of flexure observations made in 1894. 

 After setting the instrument on one of the collimators, readings were taken both 

 immediately and after an interval of two minutes, and it was found that both 

 circle and telescope had moved in the interval, the former by about 0""20 and the 

 latter by about double that quantity. Reasons have been given also for thinking 

 that the R — D discordance is definitely related to changes of this Idnd, which 

 take place soon after the instrument is set. 



Now Dr. Robinson remarks that any tremor will hasten such changes. He 

 speaks chiefly of friction, but his remark applies equally well to yielding to stress. 

 Hence any change in the tremors in the neighbourhood of an observatory will 

 probably introduce some systematic changes in the observations of zenith distance. 



The changes are of a kind which might be very difficult to explain, for they 

 might occur in spite of the fact that all the other circumstances had been pre- 

 served precisely the same. It is to be remarked also that the previous existence 

 of tremors from railways is no safeguard against a totally new interference from 

 those at present threatened. The former would naturally be intermittent ; a train 

 passes now and again, and the observation of a particular star may suffer a little 

 in the manner found by Dr. Robinson; but if the engines of the generating 

 station produce continuous tremors all the observations would be affected. 



It is impossible to do more than outline this possible source of error here in 

 default of observations of a special kind, but it seems desirable to put on record 

 these suggestions at the present moment. 



5. Speotroscojnc Observations of Solar Eclipses. 

 By Professor F. W. Dyson, M.A., F.R.S. 



6. Discussion on Radio-activity and the Internal Structure of the Earth, 

 Oj^ening remarks by Hon. R. J. Sthutt, F.B.S. 



It has long been supposed that the internal heat of the earth is merelyTa 

 remnant of the heat generated by contraction of a primseval nebula. This theory 

 appeared until lately the only possible one : it was open, however, to the objection 

 that in that case the time which could have elapsed since the earth was red-hot 

 became very short — much shorter, in fact, than the requirements of geology could 

 easily admit. Some kind of modus rivendi had been reached between geologists 

 and physicists on this matter. It may be doubted, however, whether the result 

 of the controversy was really satisfactory to either party. Two years ago the 

 suggestion was made by several writers that there might be enough radium in 

 the earth to account for the internal heat. Rutherford calculated from some 



